1 # -*- coding: utf-8 -*-
8 from librarian.parser import WLDocument
9 from librarian import XHTMLNS, ParseError
11 from lxml.etree import XMLSyntaxError, XSLTApplyError
13 ENTITY_SUBSTITUTIONS = [
22 'legacy': 'xslt/book2html.xslt',
23 'full': 'xslt/wl2html_full.xslt',
24 'partial': 'xslt/wl2html_partial.xslt'
27 def get_stylesheet(name):
28 return os.path.join(os.path.dirname(__file__), STYLESHEETS[name])
30 def substitute_entities(context, text):
31 """XPath extension function converting all entites in passed text."""
32 if isinstance(text, list):
34 for entity, substitutution in ENTITY_SUBSTITUTIONS:
35 text = text.replace(entity, substitutution)
38 # Register substitute_entities function with lxml
39 ns = etree.FunctionNamespace('http://wolnelektury.pl/functions')
40 ns['substitute_entities'] = substitute_entities
42 def transform(input, output_filename=None, is_file=True, \
43 parse_dublincore=True, stylesheet='legacy', options={}):
44 """Transforms file input_filename in XML to output_filename in XHTML."""
47 style_filename = get_stylesheet(stylesheet)
48 style = etree.parse(style_filename)
51 document = WLDocument.from_file(input, True, \
52 parse_dublincore=parse_dublincore)
54 document = WLDocument.from_string(input, True, \
55 parse_dublincore=parse_dublincore)
57 result = document.transform(style, **options)
58 del document # no longer needed large object :)
60 if etree.ETXPath('//p|//{%s}p' % str(XHTMLNS))(result) is not None:
61 add_anchors(result.getroot())
62 add_table_of_contents(result.getroot())
64 if output_filename is not None:
65 result.write(output_filename, xml_declaration=False, pretty_print=True, encoding='utf-8')
72 raise ValueError("'%s' is not a valid stylesheet.")
73 except (XMLSyntaxError, XSLTApplyError), e:
76 class Fragment(object):
77 def __init__(self, id, themes):
78 super(Fragment, self).__init__()
83 def append(self, event, element):
84 self.events.append((event, element))
86 def closed_events(self):
88 for event, element in self.events:
90 stack.append(('end', element))
95 print 'CLOSED NON-OPEN TAG:', element
98 return self.events + stack
102 for event, element in self.closed_events():
104 result.append(u'<%s %s>' % (element.tag, ' '.join('%s="%s"' % (k, v) for k, v in element.attrib.items())))
106 result.append(element.text)
108 result.append(u'</%s>' % element.tag)
110 result.append(element.tail)
112 result.append(element)
114 return ''.join(result)
116 def __unicode__(self):
117 return self.to_string()
120 def extract_fragments(input_filename):
121 """Extracts theme fragments from input_filename."""
123 closed_fragments = {}
125 for event, element in etree.iterparse(input_filename, events=('start', 'end')):
126 # Process begin and end elements
127 if element.get('class', '') in ('theme-begin', 'theme-end'):
128 if not event == 'end': continue # Process elements only once, on end event
131 if element.get('class', '') == 'theme-begin':
132 fragment = Fragment(id=element.get('fid'), themes=element.text)
135 if element.getparent().get('id', None) != 'book-text':
136 parents = [element.getparent()]
137 while parents[-1].getparent().get('id', None) != 'book-text':
138 parents.append(parents[-1].getparent())
141 for parent in parents:
142 fragment.append('start', parent)
144 open_fragments[fragment.id] = fragment
146 # Close existing fragment
149 fragment = open_fragments[element.get('fid')]
151 print '%s:closed not open fragment #%s' % (input_filename, element.get('fid'))
153 closed_fragments[fragment.id] = fragment
154 del open_fragments[fragment.id]
156 # Append element tail to lost_text (we don't want to lose any text)
158 for fragment_id in open_fragments:
159 open_fragments[fragment_id].append('text', element.tail)
162 # Process all elements except begin and end
164 # Omit annotation tags
165 if len(element.get('name', '')) or element.get('class', '') == 'annotation':
166 if event == 'end' and element.tail:
167 for fragment_id in open_fragments:
168 open_fragments[fragment_id].append('text', element.tail)
170 for fragment_id in open_fragments:
171 open_fragments[fragment_id].append(event, copy.copy(element))
173 return closed_fragments, open_fragments
176 def add_anchor(element, prefix, with_link=True, with_target=True, link_text=None):
178 if link_text is None:
180 anchor = etree.Element('a', href='#%s' % prefix)
181 anchor.set('class', 'anchor')
182 anchor.text = unicode(link_text)
184 anchor.tail = element.text
186 element.insert(0, anchor)
189 anchor_target = etree.Element('a', name='%s' % prefix)
190 anchor_target.set('class', 'target')
191 anchor_target.text = u' '
193 anchor_target.tail = element.text
195 element.insert(0, anchor_target)
198 def any_ancestor(element, test):
199 for ancestor in element.iterancestors():
205 def add_anchors(root):
207 for element in root.iterdescendants():
208 if any_ancestor(element, lambda e: e.get('class') in ('note', 'motto', 'motto_podpis', 'dedication')
209 or e.tag == 'blockquote'):
212 if element.tag == 'p' and 'verse' in element.get('class', ''):
213 if counter == 1 or counter % 5 == 0:
214 add_anchor(element, "f%d" % counter, link_text=counter)
216 elif 'paragraph' in element.get('class', ''):
217 add_anchor(element, "f%d" % counter, link_text=counter)
221 def add_table_of_contents(root):
224 for element in root.iterdescendants():
225 if element.tag in ('h2', 'h3'):
226 if any_ancestor(element, lambda e: e.get('id') in ('footnotes',) or e.get('class') in ('person-list',)):
229 if element.tag == 'h3' and len(sections) and sections[-1][1] == 'h2':
230 sections[-1][3].append((counter, element.tag, ''.join(element.xpath('text()')), []))
232 sections.append((counter, element.tag, ''.join(element.xpath('text()')), []))
233 add_anchor(element, "s%d" % counter, with_link=False)
236 toc = etree.Element('div')
238 toc_header = etree.SubElement(toc, 'h2')
239 toc_header.text = u'Spis treści'
240 toc_list = etree.SubElement(toc, 'ol')
242 for n, section, text, subsections in sections:
243 section_element = etree.SubElement(toc_list, 'li')
244 add_anchor(section_element, "s%d" % n, with_target=False, link_text=text)
247 subsection_list = etree.SubElement(section_element, 'ol')
248 for n, subsection, text, _ in subsections:
249 subsection_element = etree.SubElement(subsection_list, 'li')
250 add_anchor(subsection_element, "s%d" % n, with_target=False, link_text=text)